The vision of mental health recovery promoted by the President's New Freedom Commission portends a transformation of services for psychiatrically disabled persons, yet empirical research on recovery has not kept pace with its rhetorical appeal. For 'dual diagnosed' homeless persons, paths to recovery are even less described or understood despite the fact that they incur disproportionately high medical costs and live amidst concentrated disadvantage. Coincidentally with the recovery movement in the 1990s was the rise of 'housing first', a service approach for mentally ill homeless adults that distinctly contrasts with usual care, or 'treatment first'. Like mental health recovery, 'housing first' has received Federal and state endorsements yet is not yet widespread and remains understudied. This multi-phase qualitative longitudinal study will draw on relevant theoretical frameworks to examine mental health recovery among 'dual diagnosed homeless' adults entering residential services in New York City. Specific Aims are: 1) To identify common elements and contexts of recovery among dual diagnosed homeless adults (DDHAs) served by 'housing first' (HF) and 'treatment first' (TF) programs. This aim will be accomplished by using independent nominations criterion sampling to select individuals who manifest relatively high levels of functioning two years after entering a TF program (n=10) or a HF program (n=10). An additional 10 long-term (>5 years) HF and TF samples will be criterion-sampled and recruited. These 40 individuals will be interviewed in-depth and case study analyses used to develop one or more models of recovery to inform Aim 2; 2) To conduct a prospective qualitative study of dual diagnosed homeless adults (DDHAs) entering HF and TF programs in order to document barriers and incentives for recovery and their 'real time' sequences and contexts. Model(s) developed in Aim 1 will inform Aim 2 inquiries into individual (Aim 2a), provider (Aim 2b) and organizational (Aim 2c) factors. Aim 2a will include 4 in-depth interviews over an 18-month period with 50 new enrollees (25 each at HF and TF programs), 3 interviews with their collateral persons, and 'shadowing' observation of a sub-sample of 10 enrollees (6 days per person). Aim 2b will include 3 interviews with their case managers and Aim 2c will involve targeted ethnographic visits to the program sites (n=24 visits). Individual case-centered data collection and analyses will be used to develop an inductive model of recovery relevant to dual diagnosed adults seeking to exit homelessness. Provider- and organization-centered data collection and analyses will be used to identify barriers and incentives to recovery-oriented services emanating from the service system context. Aim 3: To recommend changes that will facilitate implementation of recovery-oriented practices (ROPs) in working with the dual diagnosed homeless population. This aim will involve integrating findings from Aims 1 and 2 to recommend organizational and provider- related changes needed to foster recovery among DDHAs. To accomplish this aim, we will conduct a preliminary study to explore whether these ROP recommendations are acceptable, appropriate and feasible in three regional settings: New York City, Philadelphia, and Washington DC. This aim is intended to take the findings 'up and out' to explore their translatability in both HF and TF programs (6 programs in 3 cities). Specifically, we will conduct 15 focus groups, one with front-line staff and another with supervisory staff at each HF and TF program and one focus group of consumers per city. Data analyses from these groups will inform a set of recommendations for ROPs that take into account diverse settings and programs. The proposed study seeks to start broadly from the consumer's perspective, expand to include significant others (collaterals, case managers and organizations) and incorporate observational data to maximize analytic power and the potential for translation and dissemination to improve 'real world' services for a severely disadvantaged population.